watchdog for rspamd + redis

Discussion in 'Feature Requests' started by elmacus, Sep 26, 2019.

  1. elmacus

    elmacus Active Member

    My restart of several servers and redis + rspamd did not start correctly, to only after manual start.
    So probably watchdog needed.
     
  2. till

    till Super Moderator Staff Member ISPConfig Developer

    Please check if Rspamd and Redis are enabled to start automatically in systemd. If this does not help, try adding a start command to rc.local. ensure that Redis is started before rspamd is started.
     
  3. elmacus

    elmacus Active Member

    redis-server.service - Advanced key-value store
    Loaded: loaded (/lib/systemd/system/redis-server.service; enabled; vendor preset: enabled)
    Active: active (running) since Wed 2019-09-25 16:16:58 CEST; 17h ago

    rspamd.service - rapid spam filtering system
    Loaded: loaded (/lib/systemd/system/rspamd.service; enabled; vendor preset: enabled)
    Active: active (running) since Wed 2019-09-25 16:16:58 CEST; 17h ago

    Im not sure on how to check what service starts first in systemd, will check my favorite search engine, but if you know, please tell.
     
  4. elmacus

    elmacus Active Member

    nano /etc/systemd/system/multi-user.target.wants/rspamd.service
    #Add redis.service last in line [Unit] After=
    After=nss-lookup.target network-online.target redis.service

    systemctl daemon-reload
    I hope this is it.
     
    till likes this.
  5. till

    till Super Moderator Staff Member ISPConfig Developer

    Please let me know if this fixes the issue, I will add it to the installation instructions then.
     
    elmacus likes this.
  6. elmacus

    elmacus Active Member

    Yes, but it will take some time since this is production servers and i cant restart normaly during office hours. (Yesterday a physical server was fenced and all VM moved and started elswhere, Proxmox 6, sigh).
     
  7. till

    till Super Moderator Staff Member ISPConfig Developer

    sure, no hurry :) I will test it here on my test systems too.
     
  8. elmacus

    elmacus Active Member

    Test machine restarted and no problem, need to check systemd logs if it worked.
     
    till likes this.
  9. elmacus

    elmacus Active Member

    Proxmox HA killed a node, so several VM had to be restarted, none worked after reboot.
    Had to manually start mysql, dovecot, redis, rspamd and restart postfix, in that order.
    So Debian 10 is not that fun yet.
    My instuctions above did not work or need more dependencies.
     
  10. till

    till Super Moderator Staff Member ISPConfig Developer

    I guess this must be related to the type of virtualization system as my Debian 10 test systems, virtualized with VMWare, start without issues and all services are up.
     
  11. elmacus

    elmacus Active Member

    Manually restarting another server now did not fail.
    So its when they crash, i need to hardstop them, then they dont come up correct.
     
  12. florian030

    florian030 Well-Known Member HowtoForge Supporter

    I don't think, that this is related to the os from the vms. Did you check the logs on the host?
     
  13. elmacus

    elmacus Active Member

    Just a feedback on this matter, Monit fixed the problem.
     
    till likes this.
  14. felan

    felan Member HowtoForge Supporter

    I am running Proxmox with Debian 10 and perfect server multiserver setup. I have had no issues with it so far.
     
  15. elmacus

    elmacus Active Member

    Did your server crash yet ? Else you dont know if you have problem.
     
  16. felan

    felan Member HowtoForge Supporter

    Ive had a crash a few days after switching to rspamd, back when the first tutorial came out. A restart of the vm and it was all working again.
     

Share This Page